I started on a slag-removal crew for a traveling industrial contractor. Two years of carrying 80-pound blast hoses through running coal plants, threading pipe that would be imminently destroyed by the online deslagging operations. Everything after is both icing on the cake and built on a foundation of relentless and joyful hard work.
From the boiler floor I went to work for Dayton Power & Light on outage planning. AES bought DP&L in 2011 and got me with the acquisition; I never left. From there: rotating-equipment engineering, planning for global and U.S. generation, Senior Manager over Global Generation Asset Management. The last four years of my corporate career I led Utilities Asset Management at AES. Reliability engineering, investment planning, traditional T&D planning (system forecasting, modeling, and analysis), and norms and standards, across two midwest utilities. That's where I authored PUCO expert testimony for the OH smart grid phase 2 program. Somewhere in there I taught myself Python because Excel stopped scaling. And somewhere in there I watched several "digital transformation" programs arrive, consume massive amounts of both labor and financial resources, then leave.
§ 01Why your enterprise software implementation didn't deliver the results you were promised
The big consultancies ship decks. The platform vendors ship dashboards. Both solve for the VP who signed the contract, not the engineer who was there at 2 AM. A tool only changes an operations organization when the operators open it the next morning.
Nobody in my control room ever reopened the eight-figure platforms.
That was the lesson. A tool that survives past the engagement has to be something an engineer at the utility can crack open, read, and change. Black-box platforms don't do that. Consulting PowerPoints don't do that. A well-documented notebook with a real model and a README does.
§ 02What I actually build
Four service areas, each delivered with our own tools and models.
Grid Advisory is utility-side modernization work: strategy, implementation oversight, regulatory support, expert testimony when the math needs to hold up under cross-examination.
Grid Reliability ports the Google SRE methodology to distribution. Per-feeder error budgets, burn-rate alerting, Weibull regime trending, and a CEMI-5 equity overlay, because SAIDI averages hide the customers who actually carry the outages.
ML & Automation is the model-building line. Outage prediction, load forecasting, hosting capacity, predictive maintenance, all trained on your data and documented so your engineers maintain it after I'm gone.
Data Center Developers is project-viability work for sites where deliverable power is the schedule risk. BTM-Optimize is the stochastic model behind a three-tier engagement: site screen, power stack model, developer package.
All of it is grounded in the same worldview: the grid is stochastic, not deterministic, and your decisions should reflect that.
§ 03How to work with me
Engagements are 6 to 16 weeks, fixed-scope. Retainer work exists but is rare; I'd rather ship a thing and go. Audits (one week, one deliverable, one flight) are the fastest way to find out if we're a fit. The first call is always a diagnostic: 30 minutes, and if I'm not the right person for what you're trying to move, I'll tell you who I'd call instead.
§ 04Faith+Works
I am a Christian of the thoroughly Reformed persuasion. The catechism I was shaped by opens with a single question and a single answer:
Q. What is the chief end of man?
A. Man's chief end is to glorify God, and to enjoy him forever. — Westminster Shorter Catechism, Q1
The execution and application of this applies to my work as well. The grid is infrastructure people live on. Hospitals, home heaters, the stove where a family makes dinner. Keeping it reliable is one of the more concrete ways I know how to love a neighbor at scale. It's why SAIDI is a currency to me and not a scorecard, and why I don't think any of this is "just" software.
— Adam